Post-treatment cervical cancer surveillance cytology and recurrence outcomes in a large integrated healthcare system.

Abstract

e17507 Background: Post-treatment cervical cancer surveillance strategies are used to facilitate timely detection of recurrence and commonly includes annual vaginal or cervical cytology, despite limited evidence of its clinical utility for detecting recurrence. This study aimed to determine whether abnormal post-treatment vaginal/cervical cytology was a risk factor contributing to cervical cancer recurrence and to identify additional independent predictors of recurrence. Methods: We conducted a retrospective cohort study using electronic health records of patients with cervical cancer between January 1, 2011, and December 31, 2020, in an integrated health system who completed at least two consecutive years of post-treatment surveillance. Recurrence rates were compared between patients with abnormal and normal cytology results using chi-square tests for categorical variables and ANOVA for continuous variables. Multivariate analysis was performed using logistic regression to identify independent predictors of recurrence and Cox proportional hazards regression with Fine and Gray competing risks modeling to estimate subdistribution hazard ratios for time to recurrence. Results: Among 365 patients, 52 (14.2%) experienced recurrence at a median of 1.9 years (IQR 1.3–3.9) post-treatment. Recurrence was associated with older age at diagnosis (median 51.5 vs 45.0 years, p<0.05), higher stage (I: 10.5%, II: 24.4%, III: 27.7%; p<0.01), and treatment type (surgery: 7.5%, concurrent chemoradiation therapy (CCRT): 23.9%, combined: 28.1%; p<0.001). Abnormal cytology was not significantly associated with recurrence (OR 1.1, 95% CI 0.6–2.0, p=0.78). In multivariate analysis, surgery followed by adjuvant chemoradiation remained a significant predictor compared with surgery alone (OR 4.4, 95% CI 1.8–10.9, p=0.001). A treatment intensity gradient from surgery alone through chemoradiation alone to combination therapy showed significant association with recurrence (p<0.001,effect size 0.26). Higher cancer stage at diagnosis and lymph node involvement were also significantly associated with recurrence (p<0.01), with an effect size of 0.19 for the former and 0.15 for the latter. Conclusions: Treatment modality was the strongest independent predictor of recurrence, reflecting higher baseline disease burden. Age, stage, and lymph node involvement define recurrence risk, not cytology.